In recent years, for example, in a radio communication system, such as a mobile communication system, or the like, data is transmitted with high efficiency due to digitization. Specifically, for example, a multilevel phase modulation scheme is sometimes used in order to transmit multiple-bit data by a single symbol. If the multilevel phase modulation scheme is used in the radio communication system, at the transmission side, it is desirable to control nonlinear distortion by linearizing the characteristic of amplification of an electrical power amplifier that amplifies a transmission signal. However, in order to enhance the linearity of the electrical power amplifier, there is a need to use an expensive device or to increase the voltage applied to the electrical power amplifier, resulting in an increase in cost or electrical power consumption.
Thus, as a technology of distortion compensation that compensates nonlinear distortion generated in the electrical power amplifier, for example, a digital nonlinear distortion compensation (digital predistortion: DPD) scheme is sometimes used. The DPD compensates nonlinear distortion by multiplying a transmission signal by a distortion compensation coefficient that has the inverse characteristic of the nonlinear distortion generated in the electrical power amplifier. If the transmission signal in which the distortion compensation coefficient is multiplied is input to the electrical power amplifier and amplified, the distortion compensation coefficient is cancelled by nonlinear distortion at the time of amplification and thus it is possible to increase the linearity of an input and an output to and from the electrical power amplifier.
Various methods are used to implement the DPD and, in general, a lookup table (LUT) method is known. The LUT method is a method that determines addresses on the basis of electrical power of an input signal and that refers to and updates the distortion compensation coefficients stored in the LUT. The distortion compensation coefficients stored in the LUT are updated on the basis of the result of comparing an input signal with the amplitude of a feedback signal that is fed back after the amplification.
Patent Document 1: Japanese Laid-open Patent Publication No. 2011-199428
Patent Document 2: Japanese Laid-open Patent Publication No. 2006-279528
Patent Document 3: Japanese Laid-open Patent Publication No. 2010-141725
Patent Document 4: Japanese Laid-open Patent Publication No. 2014-158220
However, the distortion compensation coefficients stored in the LUT is not sometimes appropriately be updated because of a change in analog characteristics caused by a change in, for example, temperature, or the like, or because of a change in input signal. In this case, distortion compensation coefficients diverge and, consequently, adjacent channel leakage electrical power or carrier electrical power may possibly be increased. Thus, after having compared the amplitude of an input signal with that of a feedback signal, if there is a difference equal to or greater than a predetermined value, it is conceivable to stop the operation of a device assuming a failure of the device. Namely, if an error between the feedback signal and input signal is great even though the feedback signal is a signal in which the nonlinear distortion generated in the input signal has been canceled by distortion compensation coefficients, it is conceivable that appropriate distortion compensation is not performed. Consequently, if an error between the input signal and the feedback signal is great, an increase in adjacent channel leakage electrical power or the like is prevented due to a stop of the operation of the device.
Furthermore, studies have been conducted on a method of, for example, monitoring the amplitude of the signal that is output from a distortion compensation circuit and stopping, if the amplitude exceeds the allowable fluctuation range, it is also conceivable to stop distortion compensation performed by using the distortion compensation coefficients. Consequently, it is possible to stop the distortion compensation performed by inappropriate distortion compensation coefficients without stopping the operation of the device.
However, there is a problem in that, although memory effect or nonlinearity is present in the electrical power amplifier, detection of an abnormality of the distortion compensation is performed without sufficiently taking into consideration the memory effect or nonlinearity. Specifically, for example, even if an influence of the signal amplified in the electrical power amplifier in the past is received due to the memory effect in the electrical power amplifier and distortion compensation has been appropriately performed, the magnitude of the error between the input signal and the feedback signal may possibly be increased. Furthermore, because the distortion compensation coefficients are coefficients that reflect the nonlinearity of the electrical power amplifier, the allowable fluctuation range varies due to the amplitude of the signal that is output from the distortion compensation circuit.
Consequently, when an abnormality of the distortion compensation is detected, it is difficult to accurately detect an abnormality of the distortion compensation by only uniformly comparing the index value, such as an error between the input signal and the feedback signal, the amplitude output from the distortion compensation circuit, or the like, with a predetermined value.